Update app.py

app.py

@@ -1,10 +1,13 @@
+"""
+Fun little experiment, written with o3. 
+"""
+
 import gradio as gr
 import torch
 import concurrent.futures
 from transformers import AutoTokenizer, AutoModelForCausalLM
 
-# Load the model and tokenizer (using GPT-2 as an example)
-model_name = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"
+model_name = "gpt2"
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 model = AutoModelForCausalLM.from_pretrained(model_name)
 
@@ -133,7 +136,7 @@ interface = gr.Interface(
         gr.Textbox(label="Min-p Sampling"),
     ],
     title="Decoding Methods Comparison",
-    description="Each decoding method's final answer is printed as soon as it is done, including custom min-p sampling."
+    description="Each decoding method's final answer is printed as soon as it is done. This uses GPT2."
 )
 
 if __name__ == "__main__":